Patrons @andrewb9830 and @wearwolf2500 let me know that the white PCB TEAC does have the "I" jumper just to the right of the drive select jumpers. I just retested the drive and indeed, dual speed! Also, @wearwolf2500 reported: "I tested the termination on my FD-55GFR the other day and it measured 2K ohms." This is instead of the normal 150 ohm resistor chip that should be on the last drive in the chain of drives on the cable.
Hey Adrian, Love the video. I found this post in an old forum about the Chinon FZ-506 jumpers. You can test it out to see if it helps. TM: closed for drive A, open for drive B DS0: open DS1: closed DS2: open DS3: open IU: open MS1: closed D-R: closed MS2: open TM are the termination resistors, which must be enabled only for the last drive on the cable, normally drive A. DS1 is the standard Drive Select number for IBM PC/compatibles. D-R selects the function of pin 34, Disk Changed if open, Drive Ready if closed. IU enables the In Use signal, which is, ironically, never actually used. Enabling both MS1 and MS2 selects automatic motor speed switching. MS2 enables it, and MS1 inverts the signal from pin 2 (or else it spins at 360rpm for DD and 300rpm for HD discs).
on that 4th Teac drive with the different circuit board there is a jumper labeled I next to one labeled LG. It's located along the bottom edge above resistors r35 and r36.
I really like these kind of videos - usually I do computery stuff while watching, like imaging disks, testing stuff, etc. - keep up your very good videos! :)
I just plain old sat here watching and sipping coffee. Thinking about my old XT computer in the other room in it's box under my bed. Thinking I should get that out, open it up, check things over, see if the clock battery leaked, try to boot it. Scary.
Hey Adrian, you might want to construct a "diagnostic" PCB from some protoboard. I made one that goes between a drive and PC, and it has LEDs showing the logic states of each line. I can also manually command lines high or low to run the drives without a PC. You can also use the drive's index pulse connected to the step input to make the head move once per rotation. Making the board gave me a really great understanding of how the interface works. Would be a cool project!
FYI the twisted cable was an invention of IBM. They figured most users were too stupid to use jumpers. Then they set all their drives to drive 1. So they could just ship a drive and plug it in. In most cases, even on the IBM's I played with in the day, you could use a straight through cable and jumper the drives correctly. This is what most CPM and other computers did.
The reason for the twist was the original PC 5150 only had a 65W (!) power supply, not enough to run both drive motors at the same time. By flipping the center pins, DS2 (Drive Select 2) was sacrificed for an extra motor select connection (for the "B:" or DS1 drive). That made copying a file from A: to B: very slow (each drive would have to spin up with each re-select). Your straight-through connection would work fine, mostly, as long as you set the drive jumpers for motor-on-with-select... although some BIOSes would cause you to experience intermittent errors caused by the drive motor not running when the BIOS asked for it.
There's a whole new step in your voice. More like "I have all day, no rush" step. Your content is great, and I'm sure we all thank you for your contributions to the community.
As opposed to using rubber belts (sadly i have one usb drive with that). I love direct drive motors, those last and unless you used the PC in a construction site or somewhere super dirty, very rarely needed any maintaining. People who belittle floppies lack the context of cassette tapes, and those that belittle cassette tapes lack the context of punch cards/ribbon type your program every time...
Especially compared to hard drives from that era. Adrian has opened up more than one HDD on his show, revealing a read/write head that had gotten stuck to its platter and then _torn off_ by the spindle motor.
Well, most of us got exited over diskette drives in the 1970s and early 80s, when tape storage was the norm for hobby computers. Plus, the techniques or solutions used can be just as fascinating as with hard disks, or even SSDs. Especially when implementing parts of it yourself (say an interface or low level read/write).
Some of the very earliest videos (assuming they’ve not been taken down since I watched) showed the basement near-empty. Also it’s kind of eerie to be like “that’s where this bench comes in later, that’s where these shelves go”. It’s pretty big but I’ve also definitely seen bigger in some American suburban basements.
Hey Adrian, I have a drive that made exactly the same sound as yours at 36:45 in the video. Turns out the problem was simply the flywheel rubbing against the surface of the desk! Prop the drive up on the edges, install it in a drive bay, or lie it on its side and it will probably work just fine 🙂
All 55GFRs are dual speed, including that 7149-u5. There is a I-jumper right to the drive select ”cross” jumper on your video. I just checked the manual and it says that when I jumper is not set, speed will be always 360 RPM. When it is placed, it will be 300rpm in low density mode and 360rpm in high density mode. Default is jumper is not placed and this is also the situation on your drive. IU-jumper is ”In Use” signal and it affects the LED.
~For the fourth TEAC drive: There is an I jumper silk mark next to the bent LG jumper location right (on camera) of the D1/2/3/4 jumpers, maybe try that for Dual speed.~ You already found it!
I am so envious that you have so many 5.25" drives and I have only one in my 386. I never use it but my life would not have meaning if I did not get to hear boot up floppy seek from both the 3.5" and 5.25" drives every time I go to play Dune 2.
Fun video, I hope you do a video like this with Commodore 1541 drives I have a flip down arm 1541 that is the kind that will have heads that physically break off. I would love to see all your tips on getting 1541 drives running and calibrated. Also, the Coleco Adam has a floppy drive, but those are cRaZy expensive.
Good video, I now know more than I ever wanted about floppy drives! Keep it up. Also happy you went full time on your channel. Living your dream. Not many of us can say that.
Way back in the 80s, I acquired a faulty 80 track drive through the Northern Suburbs Color Computer Club which I intended to use on my coco. The drive would not unload the head as the solenoid had nothing on the armature to provide a magnetic gap. The residual magnetism kept it held to the solenoid core. Bit of tape on the armature and head unloaded every time. It was a double density one, not high density. One of the members had a “bridge disk” which was a 1.2MB drive with the dual speed jumper. He was having trouble using it on a PC and I could hear the drive change speed. I figured the DS jumper (it was labelled DS) was the issue and changed it. Drive now didn’t change speed and worked fine. Congratulations on becoming a full time UA-camr, hope you have a great new career Adrian!
I literally crawled every inch of VCF Midwest trying to find a 5.25" floppy drive for an archival project and ended up having to spend $60 for a TEAC on eBay. Now I know where they all went! 🤣
I was about to write the same thing. I've wanted one for 2 years to complete a 486 but I don't feel like shipping an enormous quantity on Ebay. I always hope for a miracle.
I am still not over breaking 5 5.25 floppy drives with that cursed cleaning disk 2 years ago. I need to finally make some attempts at repairing them and I have more of them to test. Thanks for reminding my of the procedures and software to use.
Yeah, but, if you tested it in DOS, we could have seen Our Pets Records! Nice haul Adrian... Oh yeah, afterthought... You might consider putting the date you tested them on the drives! As you know, time flies... 👍
Interesting! I just working on a video now that is about an external 5.25" drive for the Amiga. It has a Teac FD 55GFR in it with a small pcb that makes it work with the Amiga. It also have a 40/80 track switch
Watched the first 7 minutes and my brain exploded. Had to rewatch a lot of it twice to get my head around the information.. Damn.. I'm truly perplexed.. ❤ But seriously I've been utterly dumbfounded by pretty much every video you've done and I have watched nearly all of them . Huge thanks and mad respect..
Watching you test those drives reminded me of when I salvaged a CD-ROM drive from a very weathered computer chassis someone had left on a vacant block, it was the last component left in the machine and the faceplate was quite yellowed. However when I stuck it into a five & a quarter inch external drive casing it started up and ran with no problems for many years, sadly I lost it in a move.
When you started - picked up all those GFRs and pondered how many of those would work - I guessed 'more than half of them will be OK.' Well technically I was right but... well WAY more than half of them. It's very nice.
53:23 I'm pretty sure I have an identical drive. I'm pleased your supporters provided the correct jumper configuration; although mine appears to be fine (unlike the computer it's in, which is currently propping up the telly shelf) I never noticed it changing speeds with 360K vs. 1.2M disks .. although the noise from the machine itself probably masked that. I archived a load of 5¼" DD & HD disks before I tried upgrading the RAM and promptly losing the original, just before testing, only to see the new RAM wouldn't let the machine POST .. D'Oh! I'll have to ask the cat. 🐈
Hi Adrian! I have two Mitsumi/ Newtronics 1.2 MB drives amongst others - one of them works extremely well and reads all marginal disks, while the other one (and several TEAC drives I have) gets broken all the times after defective disks (and cleaning does not seem to help much unfortunately). They are a bit different - D509V2 and D509V3 - but there is a pair of jumpers there named either 2S/LS or 2S/HS, and 2S is for allowing the dual speed.
I swear that the ONLY Teac drives in that model generation I ever saw fail, were actually just DOA. I know I wasn't using these in the modern age, but across the 3 companies I was working for in the 80's and 90's we sold 10's of thousands of those. Often a PC would come in for 5 or 6 major upgrades over its life, but only the Teac remained stock.
I remember in the 90ies many computers 286/386 were sold with both 1.2m and 1.4m drives but by then 1.4m got traction and many users never used the 1.2m for anything in its whole PC lifespan. But that's how the shop i worked at sold them (maybe to get rid of the inventory), until the PCs were sold only with floppy/hd and maybe cd later. Those people who did have old 5 1/4" floppies were mostly coming from XT clones and those were universally 360k. So the rare 1.2m floppy i saw was because they bought a box of blanks or the rare game program that came in those. While i have a lot of floppies, i don't think i have even one 1.2mb, or maybe i do in some game box that came with both. Actually now i remember when i switched to 286 i brought my old 360k floppy drive since ALL my older floppies were 360k, and to buy new blanks i just went straight to 1.4. Floppies were reliable in the early 90ies, unless you did something stupid like putting it next to a magnet (like a speaker) or EM source like a PSU/CRT (especially self de-gaussing CRTs which most were).
Vogons has a scanned PDF of the manual for the FZ-506 floppy drive. It has multiple different modes, and from what I can tell, it should do dual speed if you get it setup correctly. I think you were pretty close though.
According to what I've read, high density drives (1.2M & 1.44M) also used a different write bias on the read/write head. For high-density mode, the magnetic field was intensified because the high-density disks were supposed to have a magnetic coating which was less "permeable" to magnetism. This was done to reduce the likelihood of cross-talk between the closer-spaced tracks (80 tracks instead of 40 tracks). Putting double the number of tracks into the same space meant that the magnetic fields for each bit in adjacent tracks were far closer to each other than in double-density drives, and over time the magnetic fields in adjacent tracks could interfere with each other and effectively "demagnetise" both bits, if a standard double-density disk was formatted using a high-density drive. The reduced magnetic permeability of high-density disks meant that the adjacent tracks would not interfere with each other.
Geez, if that was already starting to become an issue, how the heck did we manage to get magnetic hard drives anywhere near the level of density we're at now???
@@stevethepocket Hard drives use smaller size heads than floppy drives use, which is why the platters of the HDD are sealed away where dust can't get to them, since even a speck of dust can be bigger than many bits of data and can also scratch the magnetic coating right off the surface of the disk if it gets caught under the read/write head. SSDs don't use moving parts at all, so they're generally longer lasting and a lot more tolerant of dust and dirt. Before SSDs, rotating-platter HDDs had already changed from one recording format to a different format which made the data more compact. That's another reason why so much data can be stored on a single HDD, although the main reason is simply the mechanical improvements in precision.
@@stevethepocketin addition to what melkior said about head size and formatting, they also swapped the magnetic surface medium’s materials a few times. Hard drives in the 80s were brown or black just like floppy disks and audio tape. These were metal oxides which made fairly large particle sizes. But that was fine because the heads were even bigger. Modern hard drives now use really high purity metal surfaces, (usually) vapour-deposited on a glass/ceramic platter. It has to be a really thin layer so as to isolate each section magnetically, as well as the high purity. Audio tape started to go in that direction with metal tape toward the end but that still paled in comparison to modern hard drive platter surfaces. The engineering for these teeny tiny magnetic fluxes is really impressive. The signal can be somewhat marginal compared to the very definite flux transitions on older magnetic media though, this is somewhat analogous to how lower voltage binary signalling (12V down to 5V down to 3.3V etc etc) requires lower noise and higher sensitivity 0-1 thresholds. They rely heavily on error correction in a lot of cases nowadays (at least 5-10 years ago when I was reading up on this), especially when a physical sector has been rewritten many times. The fact we can get 10-18TB hard drives today for the same price as 1-4TB drives a decade ago does impress me still. I’d really thought we were pushing the limit in 2015.
I said 11 working drives at the beginning of the video, and boy, I guessed it right! Watched the whole video and would watch much more if you upload it! ♥
When I first started working on PC's, like my first 486 model. I had one of those huge floppy drives in my system. I had many PC's after I helped a neighbor fix his computer with a BIOS chip replacement, he gave me computer parts. Then he got me many other jobs for people. who gave me their old PC after I built them a new one. I had about 10 of those wold floppy drives. I just kept them and installed them on other systems I sold. I still have one of them. I no longer have a motherboard to be able to connect it. I may send it to Adrian.
I have that Chinon and the jumpers are set exactly like you did in 52:40. The lever is the same color as the rest of the faceplate, no idea why yours is darker. I also have the Panasonic, but yours is AK3 while mine is AP3, jumpers set the same as 54:56. I have the Epson as well, inside a typical "mini tower" case of the 90ies.
My guess was a 10% failure rate, so I was pretty spot on: I expected one or two of the drives to be problematic, and you only ended up with one that has a fault. Mind you, I did not know what brand anything was before I made that guess: if you want to be more fair to us gamblers in the future, list off the brand and model numbers of the drives before you ask us to speculate on which ones will live or die.
There are more dual-speed 3.5" drives than you might expect: In Japan, a 1.2MB format for 3.5" disks (virtually identical to the 1.2MB format for 5.25" disks of the IBM AT) was common, so this needed 360RPM. There are "3-mode drives" (supporting 3 modes, hence the name) that support 720KB disks, 1.2MB disks and 1.44MB disks. In contrast to 5.25" drives, that only get one bit of "speed info" from the density select signal on the FDC cable, 3.5" HD drives have another source of information: The HD hole in the disk. On the interface side, stuff gets messy: Some 3.5" drives have "density select" as *output*, indicating the state of the HD hole sensor switch. Some 3.5" drives have "density select" as input, with high (not pulled down) indicating HD (that's compatible with the IBM AT 5.25" disk drive, and the convention used on PC drives), but these drives are sometimes incompatible with DD systems that do not pull this line low. Some 3.5" drives have "density select" as input, with high (not pulled down) indicating DD. Now enter the magic of 3-mode drives: If you have two inputs, the hole detector and the "density select" pin, you get 4 combinations. "No HD hole"/"density select low" is for 720KB mode. "HD hole"/"density select high" is for 1.44MB mode. One of the other combinations enables 1.2MB mode, increasing the drive speed. I forgot which of the two remaining combinations it is, but you usually just don't notice that this mode exists because the BIOS never sets it on PC-compatible computers.
Thank you so much for the video, Adrian) I'm always looking forward to it, and there's also such a big one) in fact, I really like when videos are long)
Hi Adrian, the manual for the Chinon drive is on a well known archive server where the link is censored by YT. The variable speed feature has three options - active high / active low / toggle...
Yeah, i got into that good habit back in the early 90ies, no exceptions. Lazyness cause trouble. He is in no rush, no clients are waiting, but i can see him doing it like that back then. I knew people who did, but i never wanted the risk. That stuff wasn't made for hot plugging, and any tiny little mistake could damage it. But i guess he'll have fun repairing his own mistake if/when it happens (magic smoke). I'm surprised he didn't use the compressed air can with the Panasonic, poor thing.
I had a teac drive with similar issues to your non working one. Seek and spin but no read. It ended up being that the top board (with front panel led) has come disconnect front the main board. That board is connected with a ribbon cable from right next to the head connectors, through the chassis and into a dual leaf socket on the bottom of the main board. It’s not super secure so it’s possible yours vibrated out over time like mine did. I reattached the cable and now the drive works great
Nicely done! Sure, it's grunt work, but grunt work is part of retrocomputing just as much as anything else. Kudos for making a video of it :) good for background racket, if nothing else -- I listened to it while doing some artwork. It also serves the purpose of reminding me -- I have a 1989 Panasonic hardware word proc, has a VERY weird 3.5in floppy drive, basically a DD single side drive, with a belt drive spindle... oddly, there's no way to replace the belt without either removing the RPM sense PCB or desoldering its cable, and now I need to realign that PCB. I've been putting it off for a while. But I also now have the nonstandard 26pin pinout for it (no, it's not FFC/slim, it's a standard desktop size drive with a 26pin pin header connector with power integrated and no separate Berg power connector) so I might just whip up a custom cable to try and use the optical method of aligning the drive, since it's got one of those spindles where you can tell by looking. I forget whose channel it was talked about that, it's a clever optical illusion thing. So maybe I'll get to that sooner rather than later.
Hi adrian, what are those cleaning sticks at 40:05 called? i tried searching for them with every relevant word i could think of and couldnt find them online, would be much nicer than qtips.
Creative Micro Designs always used the 2nd version of the TEAC mechanisms for double density and High Density with the FD-2000 but they had to alter I guess the jumpers youre referring to (I think) which gives me an idea.
I need a working floppy for my Atari ST, I have tried three from ebay. All did not work Hopefully I will learn something from this video. Congratulations on goin full rime
Great video. I always enjoy these kind of vintage repair. I digged out a broken 3.5" floppy from the closet the other day to see why it does not read disk with data. It only read formatted blank disk and it can format a disk, but it cannot be read on a working drive. I used Image Disk and test both heads and alignment and it passed with no problem. I just have no clue what went wrong with the drive. It is Mitsubishi MF355F-2494UL. Do you have any idea?
@@adriansdigitalbasement2 Well... I've experimented with that. So I may conclude it work but has some problems reading second side. But I had only one model of FDD (FD55GFR). By the way in DOS both sides worked fine
I had no problems running 1.2M floppy drive under Linux on newer 2010 PC which luckily still has floppy controller on board, can't remember which model of drive it was but most likely some TEAC. I was using it just to do some imaging with ddrescue.
@@adriansdigitalbasement2 So if you are planning to test that could you share results about how different drives behave. I use to restore old computers as well. Especoally that were made in USSR. So some soviet models appear to use geomentry that was not used anywhere. And Linus is barely the only way to export software there. It's kind of sad story because I had to left great paret of my collection when I was moving but anyway that might be helpful.
I got an Epson SD-600 a while back, but it hasn't got the PCB amongst the latch mech with the sensors on it, and over time I've traced out what it's meant to be wiring-wise (a switch, couple phototransitors & one resistor) so I could bodge something together eventually, aside from one part, the resistor that's on the board marked "R100", nobody seems to show what it is, if you could measure it's resistance, that'd be great cos I'd like to know if my drive actually works or not considering it was annoyingly expensive to acquire and turned out to be a lemon on arrival... :)
Is there any sort of repair documentation available for these drives? It seems like, given the original cost of these drives that they would have been repaired instead of simply scrapped and replaced…
The Teac high pitch noise might come from the disk clamp mechanism. AFAIK you have to take the whole thing apart to lubricate it. Mine had the same noise issue.
Since the Newtronics commodore drives are so prone to head failures, do you know if there is another mitsumi sssd (or even a DSSD) drive mechanism that could be swapped with the newtronics mechanism to salvage the newer 1541's?
None of the 5.25" closing mechanisms seemed to have the exact same twisting rod as my A1020 broken one (snapped clean off where the metal is the thinnest for the locking spacer). Can you even get such parts as spares?
That dirty Panasonic one was probably in/near smokers, that sure looks like the accumulation from tobacco smoke. It'll clean but it's going to be a bear.
Hi Adrian, the drive with the head problems at 39:22 : perhaps I'm wrong but the solder joints from the head ribbon cable look a bit suspicious, especially lines 6, 5 and 4. Just in case :)
That non-working TEAC almost sounds like mine does, I have to lift the lever just slightly to get it to run smoothly, like it's using too much force on the center or dragging somehow.
What you say about speeds might finally explain something. Back at college we had BBC Micro floppies that worked as 40 or 80 tracks but (unlike, later, on PCs) you had to flip a switch on the front to change modes - I wonder if that switch had anything to do with switching speeds.
While I have no indication that the switch does not relate to speed, I guess this is not the most likely reason. 5.25" Double Density (9 sectors at 512 byte sector size, around 17 sectors at 256 byte sector size) always uses 300 rpm drives, so as long as you don't exceed 800KB in the 80-track mode, there is no reason for the drive to spin at 360 rpm. On the other hand, if you connect an 80-track (96 tpi) drive to a computer expecting a 40-track drive to read 40 track disks, you need to use double stepping: The computer sends *one* seek pulse to get to the next track, but a 80-track drives needs *two* seek pulses to step 1/48 of an inch. So I consider it much more likely that this switch enabled double stepping on the drive level. By the way, this is why many PC/XT BIOSes that were meant to operate on 360KB floppies only work mostly perfectly with 720KB 3.5" DD drives: The BIOS does not limit the track count to 40 tracks, so a drive that is able to do 80 steps inwards if you send it 80 seek pulses can be driven with exactly the same interface code as a drive that can just do 40 tracks until it bangs against the stop. There might be an issue convincing DOS FORMAT.COM to accept that there are 80 tracks on that drive, though, because there is no BIOS support to tell DOS that this drive is not a 360KB drive, so for formatting, a 3rd party tool that does not rely on the BIOS to indicate the drive type correctly might be required.
The old 5,25" floppy drives and 1,2 MB disk´s are very good. My old drives from 1994 are working and the disk´s contain data from the 90th. All 3,5" stuff is junk.
Adrian, I suggest taking the controller board from the non-working teac Drive, and swapping it onto the one that has the stranger newer for that does not allow speed control. This is an easy way to check if it is a mechanical failure or an electronic failure between the mechanism and the controller card on the drive itself I mean. If it fails in the other Drive, then you know there's a fault on the card, and if not and therefore it works in the other Drive, that will make that drive the Dual speed, and then you can sort out the faults in the mechanism on the other one.
Does IMD test if you can actually write to the disks or just read? If only read, I'd try to write with each, then read on a different/known cood system drive.
![Adrian's Digital Basement ][](http://yt3.ggpht.com/ytc/AIdro_k7CrQRe71qN6c0Ua7A1P_I8KkxjqLa70u0zYeCvdvnKw=s48-c-k-c0x00ffffff-no-rj)
Patrons @andrewb9830 and @wearwolf2500 let me know that the white PCB TEAC does have the "I" jumper just to the right of the drive select jumpers. I just retested the drive and indeed, dual speed!
Also,
@wearwolf2500 reported: "I tested the termination on my FD-55GFR the other day and it measured 2K ohms." This is instead of the normal 150 ohm resistor chip that should be on the last drive in the chain of drives on the cable.
Hey Adrian, Love the video. I found this post in an old forum about the Chinon FZ-506 jumpers. You can test it out to see if it helps.
TM: closed for drive A, open for drive B
DS0: open
DS1: closed
DS2: open
DS3: open
IU: open
MS1: closed
D-R: closed
MS2: open
TM are the termination resistors, which must be enabled only for the last drive on the cable, normally drive A. DS1 is the standard Drive Select number for IBM PC/compatibles. D-R selects the function of pin 34, Disk Changed if open, Drive Ready if closed. IU enables the In Use signal, which is, ironically, never actually used.
Enabling both MS1 and MS2 selects automatic motor speed switching. MS2 enables it, and MS1 inverts the signal from pin 2 (or else it spins at 360rpm for DD and 300rpm for HD discs).
God damn it am i going to watch Adrian mess around with a bunch of floppy drives for an hour?
Yes. Yes I am.
you are not alone :)
on that 4th Teac drive with the different circuit board there is a jumper labeled I next to one labeled LG. It's located along the bottom edge above resistors r35 and r36.
Yes, very close to the drive select jumper
I really like these kind of videos - usually I do computery stuff while watching, like imaging disks, testing stuff, etc. - keep up your very good videos! :)
I usually watch while soldering up pcbs. Unfortunately it make it take longer, because I always have to stop and watch for a bit😂
@@joeleagles5491haha yeap same here
I just plain old sat here watching and sipping coffee. Thinking about my old XT computer in the other room in it's box under my bed. Thinking I should get that out, open it up, check things over, see if the clock battery leaked, try to boot it. Scary.
Hey Adrian, you might want to construct a "diagnostic" PCB from some protoboard. I made one that goes between a drive and PC, and it has LEDs showing the logic states of each line. I can also manually command lines high or low to run the drives without a PC. You can also use the drive's index pulse connected to the step input to make the head move once per rotation.
Making the board gave me a really great understanding of how the interface works. Would be a cool project!
FYI the twisted cable was an invention of IBM. They figured most users were too stupid to use jumpers. Then they set all their drives to drive 1. So they could just ship a drive and plug it in. In most cases, even on the IBM's I played with in the day, you could use a straight through cable and jumper the drives correctly. This is what most CPM and other computers did.
The reason for the twist was the original PC 5150 only had a 65W (!) power supply, not enough to run both drive motors at the same time. By flipping the center pins, DS2 (Drive Select 2) was sacrificed for an extra motor select connection (for the "B:" or DS1 drive). That made copying a file from A: to B: very slow (each drive would have to spin up with each re-select). Your straight-through connection would work fine, mostly, as long as you set the drive jumpers for motor-on-with-select... although some BIOSes would cause you to experience intermittent errors caused by the drive motor not running when the BIOS asked for it.
I still think its better to avoid jumpers though - it’s so small and fiddly lol😂
@@DerekLippold Hard drives and Optical drives would take revenge on you lazy types later on 🙂
There's a whole new step in your voice. More like "I have all day, no rush" step. Your content is great, and I'm sure we all thank you for your contributions to the community.
For something with mechanical moving parts, it is really amazing how resilient floppy drives are. Great content as always. Thank you.
As opposed to using rubber belts (sadly i have one usb drive with that). I love direct drive motors, those last and unless you used the PC in a construction site or somewhere super dirty, very rarely needed any maintaining. People who belittle floppies lack the context of cassette tapes, and those that belittle cassette tapes lack the context of punch cards/ribbon type your program every time...
Especially compared to hard drives from that era. Adrian has opened up more than one HDD on his show, revealing a read/write head that had gotten stuck to its platter and then _torn off_ by the spindle motor.
It amazes me that you got one defective drive out of a dozen, while I purchased 10 to get exactly 2 working ;). Amazing luck you have!
The only man alive to get excited about floppy's. I love you Adrian, never change!
Well, most of us got exited over diskette drives in the 1970s and early 80s, when tape storage was the norm for hobby computers. Plus, the techniques or solutions used can be just as fascinating as with hard disks, or even SSDs. Especially when implementing parts of it yourself (say an interface or low level read/write).
Really enjoy watching this videos while I work it's like having a friend in the workshop.
There's an "I" jumper in the bottom middle on the drive at 32:35. Could that be the one you're looking for?
Congratulations on becoming a full time UA-camR. Thanks for the brain food videos . Just how big is that basement LOL.
It's a Tardis lol.
Some of the very earliest videos (assuming they’ve not been taken down since I watched) showed the basement near-empty. Also it’s kind of eerie to be like “that’s where this bench comes in later, that’s where these shelves go”. It’s pretty big but I’ve also definitely seen bigger in some American suburban basements.
I love the vial full of jumpers. That's clever!
Hey Adrian, I have a drive that made exactly the same sound as yours at 36:45 in the video. Turns out the problem was simply the flywheel rubbing against the surface of the desk! Prop the drive up on the edges, install it in a drive bay, or lie it on its side and it will probably work just fine 🙂
1:01:30 you can “take away the terminator” but remember that it “will be back” 😂
All 55GFRs are dual speed, including that 7149-u5. There is a I-jumper right to the drive select ”cross” jumper on your video. I just checked the manual and it says that when I jumper is not set, speed will be always 360 RPM. When it is placed, it will be 300rpm in low density mode and 360rpm in high density mode. Default is jumper is not placed and this is also the situation on your drive. IU-jumper is ”In Use” signal and it affects the LED.
~For the fourth TEAC drive: There is an I jumper silk mark next to the bent LG jumper location right (on camera) of the D1/2/3/4 jumpers, maybe try that for Dual speed.~ You already found it!
I am so envious that you have so many 5.25" drives and I have only one in my 386. I never use it but my life would not have meaning if I did not get to hear boot up floppy seek from both the 3.5" and 5.25" drives every time I go to play Dune 2.
Fun video, I hope you do a video like this with Commodore 1541 drives I have a flip down arm 1541 that is the kind that will have heads that physically break off. I would love to see all your tips on getting 1541 drives running and calibrated. Also, the Coleco Adam has a floppy drive, but those are cRaZy expensive.
Good video, I now know more than I ever wanted about floppy drives! Keep it up. Also happy you went full time on your channel. Living your dream. Not many of us can say that.
Love the thumbnail for this video! You are my #1 favorite UA-camr, Adrian. Never disappointing, always relaxing and enjoyable content. Keep it up!
Way back in the 80s, I acquired a faulty 80 track drive through the Northern Suburbs Color Computer Club which I intended to use on my coco. The drive would not unload the head as the solenoid had nothing on the armature to provide a magnetic gap. The residual magnetism kept it held to the solenoid core. Bit of tape on the armature and head unloaded every time. It was a double density one, not high density. One of the members had a “bridge disk” which was a 1.2MB drive with the dual speed jumper. He was having trouble using it on a PC and I could hear the drive change speed. I figured the DS jumper (it was labelled DS) was the issue and changed it. Drive now didn’t change speed and worked fine. Congratulations on becoming a full time UA-camr, hope you have a great new career Adrian!
I literally crawled every inch of VCF Midwest trying to find a 5.25" floppy drive for an archival project and ended up having to spend $60 for a TEAC on eBay. Now I know where they all went! 🤣
I was about to write the same thing.
I've wanted one for 2 years to complete a 486 but I don't feel like shipping an enormous quantity on Ebay. I always hope for a miracle.
I am still not over breaking 5 5.25 floppy drives with that cursed cleaning disk 2 years ago. I need to finally make some attempts at repairing them and I have more of them to test. Thanks for reminding my of the procedures and software to use.
That drive checking program was really interesting. Quite informative actually. I never messed around with my floppy drives like that before.
Yeah, but, if you tested it in DOS, we could have seen Our Pets Records! Nice haul Adrian...
Oh yeah, afterthought... You might consider putting the date you tested them on the drives! As you know, time flies... 👍
Interesting! I just working on a video now that is about an external 5.25" drive for the Amiga. It has a Teac FD 55GFR in it with a small pcb that makes it work with the Amiga. It also have a 40/80 track switch
Watched this while washing dishes. Disk drives more interesting than dishes.
Watched the first 7 minutes and my brain exploded. Had to rewatch a lot of it twice to get my head around the information.. Damn.. I'm truly perplexed.. ❤
But seriously I've been utterly dumbfounded by pretty much every video you've done and I have watched nearly all of them . Huge thanks and mad respect..
Watching you test those drives reminded me of when I salvaged a CD-ROM drive from a very weathered computer chassis someone had left on a vacant block, it was the last component left in the machine and the faceplate was quite yellowed. However when I stuck it into a five & a quarter inch external drive casing it started up and ran with no problems for many years, sadly I lost it in a move.
When you started - picked up all those GFRs and pondered how many of those would work - I guessed 'more than half of them will be OK.' Well technically I was right but... well WAY more than half of them. It's very nice.
53:23 I'm pretty sure I have an identical drive. I'm pleased your supporters provided the correct jumper configuration; although mine appears to be fine (unlike the computer it's in, which is currently propping up the telly shelf) I never noticed it changing speeds with 360K vs. 1.2M disks .. although the noise from the machine itself probably masked that. I archived a load of 5¼" DD & HD disks before I tried upgrading the RAM and promptly losing the original, just before testing, only to see the new RAM wouldn't let the machine POST .. D'Oh! I'll have to ask the cat. 🐈
Hi Adrian! I have two Mitsumi/ Newtronics 1.2 MB drives amongst others - one of them works extremely well and reads all marginal disks, while the other one (and several TEAC drives I have) gets broken all the times after defective disks (and cleaning does not seem to help much unfortunately). They are a bit different - D509V2 and D509V3 - but there is a pair of jumpers there named either 2S/LS or 2S/HS, and 2S is for allowing the dual speed.
6:50 love hearing Adrian's diatribes about dual speed floppy drives!
Oh I’m happy I watched this the first time around - will certainly save me some time when I get three drives next week :)
Thanks, Adrian! 🎉
32:09 Connect I jumper to the right of drive select cross jumper to enable dual speed mode.
41:49 - I'm guessing that's where it landed when it was thrown into the dumpster?
40:17 the blue part is likely a Ceramic Resonator (CR) to supply a stable frequency to the drive chip. The gunk is glue to keep it in place.
I swear that the ONLY Teac drives in that model generation I ever saw fail, were actually just DOA. I know I wasn't using these in the modern age, but across the 3 companies I was working for in the 80's and 90's we sold 10's of thousands of those. Often a PC would come in for 5 or 6 major upgrades over its life, but only the Teac remained stock.
I remember in the 90ies many computers 286/386 were sold with both 1.2m and 1.4m drives but by then 1.4m got traction and many users never used the 1.2m for anything in its whole PC lifespan. But that's how the shop i worked at sold them (maybe to get rid of the inventory), until the PCs were sold only with floppy/hd and maybe cd later. Those people who did have old 5 1/4" floppies were mostly coming from XT clones and those were universally 360k. So the rare 1.2m floppy i saw was because they bought a box of blanks or the rare game program that came in those. While i have a lot of floppies, i don't think i have even one 1.2mb, or maybe i do in some game box that came with both. Actually now i remember when i switched to 286 i brought my old 360k floppy drive since ALL my older floppies were 360k, and to buy new blanks i just went straight to 1.4. Floppies were reliable in the early 90ies, unless you did something stupid like putting it next to a magnet (like a speaker) or EM source like a PSU/CRT (especially self de-gaussing CRTs which most were).
Vogons has a scanned PDF of the manual for the FZ-506 floppy drive. It has multiple different modes, and from what I can tell, it should do dual speed if you get it setup correctly. I think you were pretty close though.
His mistake was closing MS1 (invert), only MS2 needs to be closed to enable dual speed.
32:41 there is the I jumper near R35. So it may support dual speed.
Great video! I think you missed the I -jumper on the 4th Teac drive with white PCB if you want to give it another try
Great Video Adrian! I remember working on PC's when these things were brand new!
According to what I've read, high density drives (1.2M & 1.44M) also used a different write bias on the read/write head. For high-density mode, the magnetic field was intensified because the high-density disks were supposed to have a magnetic coating which was less "permeable" to magnetism. This was done to reduce the likelihood of cross-talk between the closer-spaced tracks (80 tracks instead of 40 tracks).
Putting double the number of tracks into the same space meant that the magnetic fields for each bit in adjacent tracks were far closer to each other than in double-density drives, and over time the magnetic fields in adjacent tracks could interfere with each other and effectively "demagnetise" both bits, if a standard double-density disk was formatted using a high-density drive. The reduced magnetic permeability of high-density disks meant that the adjacent tracks would not interfere with each other.
Geez, if that was already starting to become an issue, how the heck did we manage to get magnetic hard drives anywhere near the level of density we're at now???
@@stevethepocket Hard drives use smaller size heads than floppy drives use, which is why the platters of the HDD are sealed away where dust can't get to them, since even a speck of dust can be bigger than many bits of data and can also scratch the magnetic coating right off the surface of the disk if it gets caught under the read/write head.
SSDs don't use moving parts at all, so they're generally longer lasting and a lot more tolerant of dust and dirt.
Before SSDs, rotating-platter HDDs had already changed from one recording format to a different format which made the data more compact. That's another reason why so much data can be stored on a single HDD, although the main reason is simply the mechanical improvements in precision.
@@stevethepocketin addition to what melkior said about head size and formatting, they also swapped the magnetic surface medium’s materials a few times.
Hard drives in the 80s were brown or black just like floppy disks and audio tape. These were metal oxides which made fairly large particle sizes. But that was fine because the heads were even bigger. Modern hard drives now use really high purity metal surfaces, (usually) vapour-deposited on a glass/ceramic platter. It has to be a really thin layer so as to isolate each section magnetically, as well as the high purity.
Audio tape started to go in that direction with metal tape toward the end but that still paled in comparison to modern hard drive platter surfaces. The engineering for these teeny tiny magnetic fluxes is really impressive.
The signal can be somewhat marginal compared to the very definite flux transitions on older magnetic media though, this is somewhat analogous to how lower voltage binary signalling (12V down to 5V down to 3.3V etc etc) requires lower noise and higher sensitivity 0-1 thresholds. They rely heavily on error correction in a lot of cases nowadays (at least 5-10 years ago when I was reading up on this), especially when a physical sector has been rewritten many times.
The fact we can get 10-18TB hard drives today for the same price as 1-4TB drives a decade ago does impress me still. I’d really thought we were pushing the limit in 2015.
I said 11 working drives at the beginning of the video, and boy, I guessed it right! Watched the whole video and would watch much more if you upload it! ♥
57:03 The cover does not read Newtronics but Mitsumi. Was that brand a subsidiary of Mitsumi then?
36:45 - It sounds like 'tape squeal' - A similar thing happens with audio tape if it's too dry.
When I first started working on PC's, like my first 486 model. I had one of those huge floppy drives in my system. I had many PC's after I helped a neighbor fix his computer with a BIOS chip replacement, he gave me computer parts. Then he got me many other jobs for people. who gave me their old PC after I built them a new one. I had about 10 of those wold floppy drives. I just kept them and installed them on other systems I sold. I still have one of them. I no longer have a motherboard to be able to connect it. I may send it to Adrian.
And it just keeps getting better!!!
I have that Chinon and the jumpers are set exactly like you did in 52:40.
The lever is the same color as the rest of the faceplate, no idea why yours is darker.
I also have the Panasonic, but yours is AK3 while mine is AP3, jumpers set the same as 54:56.
I have the Epson as well, inside a typical "mini tower" case of the 90ies.
My guess was a 10% failure rate, so I was pretty spot on: I expected one or two of the drives to be problematic, and you only ended up with one that has a fault. Mind you, I did not know what brand anything was before I made that guess: if you want to be more fair to us gamblers in the future, list off the brand and model numbers of the drives before you ask us to speculate on which ones will live or die.
Love your new camera. Details really pop now. :)
There are more dual-speed 3.5" drives than you might expect: In Japan, a 1.2MB format for 3.5" disks (virtually identical to the 1.2MB format for 5.25" disks of the IBM AT) was common, so this needed 360RPM. There are "3-mode drives" (supporting 3 modes, hence the name) that support 720KB disks, 1.2MB disks and 1.44MB disks. In contrast to 5.25" drives, that only get one bit of "speed info" from the density select signal on the FDC cable, 3.5" HD drives have another source of information: The HD hole in the disk. On the interface side, stuff gets messy: Some 3.5" drives have "density select" as *output*, indicating the state of the HD hole sensor switch. Some 3.5" drives have "density select" as input, with high (not pulled down) indicating HD (that's compatible with the IBM AT 5.25" disk drive, and the convention used on PC drives), but these drives are sometimes incompatible with DD systems that do not pull this line low. Some 3.5" drives have "density select" as input, with high (not pulled down) indicating DD.
Now enter the magic of 3-mode drives: If you have two inputs, the hole detector and the "density select" pin, you get 4 combinations. "No HD hole"/"density select low" is for 720KB mode. "HD hole"/"density select high" is for 1.44MB mode. One of the other combinations enables 1.2MB mode, increasing the drive speed. I forgot which of the two remaining combinations it is, but you usually just don't notice that this mode exists because the BIOS never sets it on PC-compatible computers.
Thank you so much for the video, Adrian) I'm always looking forward to it, and there's also such a big one) in fact, I really like when videos are long)
Hi Adrian,
the manual for the Chinon drive is on a well known archive server where the link is censored by YT. The variable speed feature has three options - active high / active low / toggle...
23:06 - Adrian, it's GOOD that you shut the power off!
'Hot-swapping' vintage components (some containing Unobtanium), is asking for trouble!
Yeah, i got into that good habit back in the early 90ies, no exceptions. Lazyness cause trouble. He is in no rush, no clients are waiting, but i can see him doing it like that back then. I knew people who did, but i never wanted the risk. That stuff wasn't made for hot plugging, and any tiny little mistake could damage it. But i guess he'll have fun repairing his own mistake if/when it happens (magic smoke). I'm surprised he didn't use the compressed air can with the Panasonic, poor thing.
I had a teac drive with similar issues to your non working one. Seek and spin but no read. It ended up being that the top board (with front panel led) has come disconnect front the main board. That board is connected with a ribbon cable from right next to the head connectors, through the chassis and into a dual leaf socket on the bottom of the main board. It’s not super secure so it’s possible yours vibrated out over time like mine did. I reattached the cable and now the drive works great
These old drives are pretty reliable. I think all drives should work.
Thanks. Hopefully you can fix the one that didn’t work. Regards /D
Nicely done! Sure, it's grunt work, but grunt work is part of retrocomputing just as much as anything else. Kudos for making a video of it :) good for background racket, if nothing else -- I listened to it while doing some artwork.
It also serves the purpose of reminding me -- I have a 1989 Panasonic hardware word proc, has a VERY weird 3.5in floppy drive, basically a DD single side drive, with a belt drive spindle... oddly, there's no way to replace the belt without either removing the RPM sense PCB or desoldering its cable, and now I need to realign that PCB. I've been putting it off for a while. But I also now have the nonstandard 26pin pinout for it (no, it's not FFC/slim, it's a standard desktop size drive with a 26pin pin header connector with power integrated and no separate Berg power connector) so I might just whip up a custom cable to try and use the optical method of aligning the drive, since it's got one of those spindles where you can tell by looking. I forget whose channel it was talked about that, it's a clever optical illusion thing. So maybe I'll get to that sooner rather than later.
Hi adrian, what are those cleaning sticks at 40:05 called? i tried searching for them with every relevant word i could think of and couldnt find them online, would be much nicer than qtips.
I'm not totally sure -- a viewer sent them in a while back. Something he found in the UK I think used for cleaning VCR heads.
DigiKey has them #473-1183-ND aka MG Chemicals brand #810D-15 described as "Swab, Double Head Synthetic Chamois Tip 5.00" (127.0mm) 15 PC"
THIS is the content I live for lol😊
I can see Adrian prepping to build a Floppotron out of those drives, lol. Have it play Prelude in D Major by Bach.
Creative Micro Designs always used the 2nd version of the TEAC mechanisms for double density and High Density with the FD-2000 but they had to alter I guess the jumpers youre referring to (I think) which gives me an idea.
I need a working floppy for my Atari ST, I have tried three from ebay. All did not work Hopefully I will learn something from this video. Congratulations on goin full rime
Very good video. I have only one 5,25 inch drive and I have to test it but the problem begins with obtaining floppy disks (at least in my country)
From my huge collection of 3.5" drives 99% of them have a solder jumper to enable the dual speed "3-mode" operation
Great video. I always enjoy these kind of vintage repair. I digged out a broken 3.5" floppy from the closet the other day to see why it does not read disk with data. It only read formatted blank disk and it can format a disk, but it cannot be read on a working drive. I used Image Disk and test both heads and alignment and it passed with no problem. I just have no clue what went wrong with the drive. It is Mitsubishi MF355F-2494UL. Do you have any idea?
Adrian Have you tried using 5.25 drives in Linux?
Nope, does it work? I'm not sure I even have a machine with a floppy interface that can run modern linux
@@adriansdigitalbasement2 Well... I've experimented with that. So I may conclude it work but has some problems reading second side. But I had only one model of FDD (FD55GFR). By the way in DOS both sides worked fine
I had no problems running 1.2M floppy drive under Linux on newer 2010 PC which luckily still has floppy controller on board, can't remember which model of drive it was but most likely some TEAC. I was using it just to do some imaging with ddrescue.
@@snap_oversteer Thank you
@@adriansdigitalbasement2 So if you are planning to test that could you share results about how different drives behave. I use to restore old computers as well. Especoally that were made in USSR. So some soviet models appear to use geomentry that was not used anywhere. And Linus is barely the only way to export software there. It's kind of sad story because I had to left great paret of my collection when I was moving but anyway that might be helpful.
I got an Epson SD-600 a while back, but it hasn't got the PCB amongst the latch mech with the sensors on it, and over time I've traced out what it's meant to be wiring-wise (a switch, couple phototransitors & one resistor) so I could bodge something together eventually, aside from one part, the resistor that's on the board marked "R100", nobody seems to show what it is, if you could measure it's resistance, that'd be great cos I'd like to know if my drive actually works or not considering it was annoyingly expensive to acquire and turned out to be a lemon on arrival... :)
I love the thumbnail! Saucy!
Is there any sort of repair documentation available for these drives? It seems like, given the original cost of these drives that they would have been repaired instead of simply scrapped and replaced…
On the different Teac drive, there's a jumper labeled 'I' right next to the D1 jumper.
I say 7 out of 10 will be working. 70%!
Hi Adrian. I wonder if one of those drives can replace the drives in a commodore 1541c? I have two of them and one of them has a dead read write head.
There is a link to the manual of the 506 over on vogons Chinon FZ-506 (96TPI, Double Sides, 1.6MB/1MB Switchable)
I predict 10 of the 12 will work. Although i'm betting some will need adjustment, and all will need cleaning and lubrication.
@32.17 the "I" jumper directly to the right of the drive select jumpers.
The Teac high pitch noise might come from the disk clamp mechanism. AFAIK you have to take the whole thing apart to lubricate it. Mine had the same noise issue.
Since the Newtronics commodore drives are so prone to head failures, do you know if there is another mitsumi sssd (or even a DSSD) drive mechanism that could be swapped with the newtronics mechanism to salvage the newer 1541's?
None of the 5.25" closing mechanisms seemed to have the exact same twisting rod as my A1020 broken one (snapped clean off where the metal is the thinnest for the locking spacer). Can you even get such parts as spares?
The first bagged Tiac drive seems to have an "I" jumper, right next to the "LG" jumper.
Small forceps (I like the ones with curved jaws) work great for jumpers, especially hard to reach and/or smaller/low profile ones.
After getting several 5.25” drives working for a couple computers, I remember why we ditched 5.25” drives.
a very funky analog instrument
Jumpers for D509V. It does do double speed it seems.
That dirty Panasonic one was probably in/near smokers, that sure looks like the accumulation from tobacco smoke. It'll clean but it's going to be a bear.
Hi Adrian, the drive with the head problems at 39:22 : perhaps I'm wrong but the solder joints from the head ribbon cable look a bit suspicious, especially lines 6, 5 and 4. Just in case :)
That non-working TEAC almost sounds like mine does, I have to lift the lever just slightly to get it to run smoothly, like it's using too much force on the center or dragging somehow.
What you say about speeds might finally explain something. Back at college we had BBC Micro floppies that worked as 40 or 80 tracks but (unlike, later, on PCs) you had to flip a switch on the front to change modes - I wonder if that switch had anything to do with switching speeds.
While I have no indication that the switch does not relate to speed, I guess this is not the most likely reason. 5.25" Double Density (9 sectors at 512 byte sector size, around 17 sectors at 256 byte sector size) always uses 300 rpm drives, so as long as you don't exceed 800KB in the 80-track mode, there is no reason for the drive to spin at 360 rpm.
On the other hand, if you connect an 80-track (96 tpi) drive to a computer expecting a 40-track drive to read 40 track disks, you need to use double stepping: The computer sends *one* seek pulse to get to the next track, but a 80-track drives needs *two* seek pulses to step 1/48 of an inch. So I consider it much more likely that this switch enabled double stepping on the drive level.
By the way, this is why many PC/XT BIOSes that were meant to operate on 360KB floppies only work mostly perfectly with 720KB 3.5" DD drives: The BIOS does not limit the track count to 40 tracks, so a drive that is able to do 80 steps inwards if you send it 80 seek pulses can be driven with exactly the same interface code as a drive that can just do 40 tracks until it bangs against the stop. There might be an issue convincing DOS FORMAT.COM to accept that there are 80 tracks on that drive, though, because there is no BIOS support to tell DOS that this drive is not a 360KB drive, so for formatting, a 3rd party tool that does not rely on the BIOS to indicate the drive type correctly might be required.
That last teac drive you check, the noise to me sounded like a bad bearing
The old 5,25" floppy drives and 1,2 MB disk´s are very good. My old drives from 1994 are working and the disk´s contain data from the 90th. All 3,5" stuff is junk.
Adrian, I suggest taking the controller board from the non-working teac Drive, and swapping it onto the one that has the stranger newer for that does not allow speed control. This is an easy way to check if it is a mechanical failure or an electronic failure between the mechanism and the controller card on the drive itself I mean. If it fails in the other Drive, then you know there's a fault on the card, and if not and therefore it works in the other Drive, that will make that drive the Dual speed, and then you can sort out the faults in the mechanism on the other one.
From one Adrian to another, do you know where I can get a 720k only 3.5? Can’t seem to find one anywhere.
Adrian, have you ever encountered a 5.25 FD double density with 80 tracks and formats to 720K? I have two rare floppy drives like this.
12 drives? RAID!!
Today im going to test a bunch of floppy drives. Its a good day.
Does IMD test if you can actually write to the disks or just read? If only read, I'd try to write with each, then read on a different/known cood system drive.
You'll have to archive that pet disk, Adrian. Can't find it on the 'net anywhere! ;D